Drug interactions in at-risk emergency department patients

Objectives: Many emergency department (ED) patients are at risk for drug interactions (DIs) because they are elders, and/or they have chronic illnesses requiring treatment with multiple medications. In the ED, medications may be added to complex treatment regimens without the benefit of screening for DIs. Emergency physicians may therefore cause DIs, or miss the opportunity to intervene against a pre-existing DI. Prior studies are contradictory regarding whether DIs are more likely to be due to medications administered or prescribed in the ED or medications prescribed elsewhere. Screening for DIs using computer software, such as that done by retail pharmacies, is now a standard of practice, and is done more frequently than when these other DI studies were reported during the previous decade. The authors monitored DIs among a focused, at-risk outpatient ED population, to test the hypothesis that ED-induced DIs have become the most common DIs in this population-at-risk. Methods: A retrospective convenience sample of 200 at-risk patients seen at a tertiary teaching hospital on selected dates of service was analyzed. Eligible patients were ED outpatients aged 60 years or more taking three or more medications, or any age taking five or more medications. Micromedex Drug-Reax software identified DIs. DIs had to represent "major" or "minor" severity, and have "excellent" or "good" literature documentation, to be scored as positive. Pre-existing versus ED-induced DIs were compared by chi-square. DIs were stratified by patient age and by number and type of medications taken. Results: The 200 outpatients (125 female, 75 male) had a mean age (+/- SD) of 64.5 (+/- 17.6) years and were taking an average of 7.2 medications. Seventy-nine pre-existing DIs were noted, occurring in 50 of the 200 patients studied. One hundred forty patients received a new medication during their ED visit. Seven new DIs (one in each of seven patients) occurred among these 140 patients. DIs were less frequently caused by medications added during the ED visit (chi(2) = 22.2, p < 0.001). Digoxin and warfarin were the sources of the greatest number of DIs. Conclusions: ED outpatients in the at-risk group frequently present with preexisting DIs. Medications initiated in the ED are a less frequent cause of DI in this group. Medication screening during an ED visit could complement the role of outpatient pharmacies and potentially improve ED patient safety. DIs are most frequently due to digoxin and warfarin in these patients.